if(T == T){
  #mar1 = c(4, 4.5, 4, 0.5);mar2 = c(4, 0.5, 4, 4)
  mar1 = c(7, 4.5, 4, 0.5);mar2 = c(7, 0.5, 4, 4)
  my_cex <- 1.3
  my_cex_lab <- 1.85
  my_cex_title <- 2.1
  scale_factor <- 100000
  par(mfrow = c(1,2))
  par(mar = mar1 ) 
  my_data_with_campaign_finance_within_caliper[,"census_HousingPriceGainScore_inflationAdj"] <- f2n(my_data_with_campaign_finance_within_caliper[,"census_HousingPriceGainScore"]) * (1-inflation_rate)
  my_data[,"census_HousingPriceGainScore_inflationAdj"] <- f2n(my_data[,"census_HousingPriceGainScore"]) * (1-inflation_rate)
  y_use <-  tapply(f2n(my_data_with_campaign_finance_within_caliper[,"census_HousingPriceGainScore_inflationAdj"]), dist_quantiles, function(x) mean(x, na.rm = T) )
  sd_use <-  tapply(f2n(my_data_with_campaign_finance_within_caliper[,"census_HousingPriceGainScore_inflationAdj"]), dist_quantiles, function(x) se(x) )
  y_use <- scale_factor * y_use
  sd_use <- scale_factor * sd_use
  #x_use <- seq(0, 20, length.out = length(y_use))
  x_use <- tapply(my_data_with_campaign_finance_within_caliper$EZPlaza_TravelDistOptimizedDist_Total_Leng, dist_quantiles, mean)
  my_lowess_control <- loess.smooth(x = f2n(my_data_with_campaign_finance_within_caliper$minNetworkOptimizedDistance_Distance_nonEZ_hwy), 
                            y = scale_factor * f2n(my_data_with_campaign_finance_within_caliper[,"census_HousingPriceGainScore_inflationAdj"]), 
                            degree = 2)
  my_lowess <- loess.smooth(x = my_data_with_campaign_finance_within_caliper$EZPlaza_TravelDistOptimizedDist_Total_Leng, 
                            y = scale_factor * f2n(my_data_with_campaign_finance_within_caliper[,"census_HousingPriceGainScore_inflationAdj"]), 
                            degree = 2)
  my_lowess_x <- my_lowess$x
  my_lowess_y <- my_lowess$y
  median_change_x <- median(f2n(my_data[!my_data$STATEFP10 %in% 39, "EZPlaza_TravelDistOptimizedDist_Total_Leng"]), na.rm = T)
  median_change_y <- median(scale_factor * f2n(my_data[!my_data$STATEFP10 %in% 39, "census_HousingPriceGainScore_inflationAdj"]), na.rm = T)
  plot(x_use, y_use, 
       main = "Change in Home Price", 
       pch = "T", 
       cex = my_cex, 
       ylim = c(min(min(y_use - 2 * sd_use)), max(y_use + 2 * sd_use)), 
       axes=FALSE, ann=FALSE)
  axis(1, cex = my_cex); axis(2, cex = my_cex)
  mtext("Miles from Intervention Site\n (E-ZPass [T] or Control [C])", side=1, line=5, cex = my_cex_lab)
  mtext("Average Change in Home Price, in USD", side=2, line=3, cex = my_cex_lab)
  title(main = "Change in Home Price", cex.main = my_cex_title)
  box()
  
  points(x = my_lowess_x, 
         y = my_lowess_y, 
         cex = my_cex, 
         type = "l")
  points(x = my_lowess_control$x, 
         y = my_lowess_control$y, 
         cex = my_cex, 
         type = "b",pch="C",lty=2,col="Gray")
  segments(x0 = x_use, 
           y0 = y_use - 2 * sd_use, 
           x1 = x_use, 
           y1 = y_use + 2 * sd_use, 
           cex = my_cex, 
           col = "gray")
  #abline(v = median_change_x, lty = 2) 
  #abline(h = median_change_y, lty = 2 ) 
  #text(x = median_change_x - 5, y = min(y_use - 1.5 * sd_use), labels = "Median Dist. to E-ZPass")
  #text(x = threshold - 8,y = median_change_y + 6500, labels = "Median Change in Home Price")
  #arrows(x0=12, y0=76000, x1=13.5, y1=79000, length = 0.10, angle = 30)
  #arrows(x0=18, y0=median_change_y + 5000, x1=17, y1=median_change_y+500, length = 0.10, angle = 30)
  
  
  par(mar=mar2 ) 
  scale_factor <- 100
  y_use <- scale_factor * tapply(f2n(my_data_with_campaign_finance_within_caliper[,"DemVoteChange2000_to_2004"]), dist_quantiles, function(x) mean(x, na.rm = T) )
  sd_use <-  scale_factor * tapply(f2n(my_data_with_campaign_finance_within_caliper[,"DemVoteChange2000_to_2004"]), dist_quantiles, function(x) se(x) )
  x_use <- tapply(my_data_with_campaign_finance_within_caliper$EZPlaza_TravelDistOptimizedDist_Total_Leng, dist_quantiles, mean)
  my_lowess <- loess.smooth(x = my_data_with_campaign_finance_within_caliper$EZPlaza_TravelDistOptimizedDist_Total_Leng, 
                            y = scale_factor * f2n(my_data_with_campaign_finance_within_caliper[,"DemVoteChange2000_to_2004"]), 
                            degree = 2)
  my_lowess_control <- loess.smooth(x = f2n(my_data_with_campaign_finance_within_caliper$minNetworkOptimizedDistance_Distance_nonEZ_hwy), 
                            y = scale_factor * f2n(my_data_with_campaign_finance_within_caliper[,"DemVoteChange2000_to_2004"]), 
                            degree = 2)
  my_lowess_x <- my_lowess$x
  my_lowess_y <- my_lowess$y
  median_change_x <- median(f2n(my_data[!my_data$STATEFP10 %in% 39, "EZPlaza_TravelDistOptimizedDist_Total_Leng"]), na.rm = T)
  median_change_y <- median(scale_factor * f2n(my_data[!my_data$STATEFP10 %in% 39, "DemVoteChange2000_to_2004"]), na.rm = T)
  plot( x_use, y_use, 
        pch = "T", 
        cex = my_cex, 
        ylim = c(min(0, min(y_use - 2 * sd_use)), max(y_use + 2 * sd_use)), 
        axes=FALSE, ann=FALSE)
  axis(1, cex = my_cex); axis(4, las=1, cex = my_cex)
  mtext("Miles from Intervention Site\n (E-ZPass [T] or Control [C])", side=1, line=5, cex = my_cex_lab)
  title(main = "Change in Dem. Vote Share", cex.main = my_cex_title)
  p <- par('usr')
  text(p[2]+2.5, mean(p[3:4]), labels = "Average Change in Dem. Vote Share", 
       xpd = NA, srt = -90, cex = my_cex_lab)
  box()
  points(x = my_lowess_x, 
         y = my_lowess_y, 
         type = "l")
  points(x = my_lowess_control$x, 
         y = my_lowess_control$y, 
         type = "b",col="gray", pch ="C")
  segments(x0 = x_use, 
           y0 = y_use - 2 * sd_use, 
           x1 = x_use, 
           y1 = y_use + 2 * sd_use, 
           col = "gray")
  #abline(v = median_change_x, lty = 2) 
  #abline(h = median_change_y, lty = 2 ) 
  #text(x = median_change_x - 5, y = max(y_use + 1.5 * sd_use + 0.05), labels = "Median Dist. to E-ZPass")
  #text(x = threshold - 8,y = median_change_y - 0.3, labels = "Median Change in Dem. Vote Share")
  #arrows(x0=12, y0=-0.20, x1=median_change_x - 0.3, y1= -0.40, length = 0.10, angle = 30)
  #arrows(x0=20, y0=median_change_y - 0.15, x1=18, y1=median_change_y - 0.03, length = 0.10, angle = 30)
  par(mfrow = c(1, 1) ) 
}
